CN110988877B - A Spaceborne High-Resolution SAR Large Squint Doppler Dewinding Method - Google Patents

Abstract

The invention discloses a satellite-borne high-resolution SAR large squint Doppler decoiling method which comprises the steps of firstly, carrying out deformation processing on satellite-borne high-resolution SAR large squint mode two-dimensional spectrum data by using cyclic shift operation; then carrying out support domain expansion processing on the deformed data; and finally, performing two-dimensional spectrum recovery processing on the data after the support domain expansion by using cyclic shift operation to realize the deconvolution of the Doppler domain of the two-dimensional spectrum data. The invention realizes the removal of the two-dimensional spectrum Doppler domain winding phenomenon, and solves the problems of energy leakage and a large number of false target conditions caused by Doppler domain spectrum winding. The decoiling method realizes frequency spectrum processing by using cyclic shift operation, avoids calculating echo data, and has very high processing efficiency and application efficiency.

Description

A Spaceborne High-Resolution SAR Large Squint Doppler Dewinding Method

technical field

The invention belongs to the field of signal processing, and relates to a spaceborne high-resolution SAR large squint Doppler dewinding method.

Background technique

In order to meet the flexible observation requirements of high-resolution SAR payloads and realize the application efficiency of target multi-angle information fusion and multi-regional rapid observation, high-resolution spaceborne SAR needs to have the ability to work under large oblique angles. Existing spaceborne SARs all work under the side-looking condition, with limited flexibility. Driven by application requirements, the high-resolution large squint working mode will be the development direction of the next-generation spaceborne SAR. However, in the wide-band wide-squinting observation mode, different from the traditional side-looking observation, the target signal will show some new characteristics. The most significant signal characteristic change is that the signal spectrum will appear wrapping in the Doppler domain. The existing spaceborne SAR signal processing algorithms are mainly designed based on the side-view observation mode. Since the signal in the side-looking mode of the spaceborne SAR does not appear the phenomenon of spectral Doppler domain wrapping, the existing spaceborne SAR algorithms do not Considering the Doppler wrapping problem, it does not have the ability to deal with the large squint observation mode, and lacks an effective means to remove the Doppler wrapping phenomenon. If the spectral Doppler warping removal process is not performed, it will lead to serious energy leakage and a large number of false targets in the large squint observation mode, which will seriously affect the image quality and target interpretation.

SUMMARY OF THE INVENTION

The technical problem solved by the present invention is: to overcome the deficiencies of the prior art, to provide a spaceborne high-resolution SAR large squint Doppler dewinding method, which realizes the removal of the two-dimensional spectral Doppler domain winding phenomenon, and solves the problem of The situation of energy leakage and a large number of false targets caused by spectral wrapping in the Doppler domain is avoided, and the method avoids the calculation of echo data, and has very high processing efficiency and application efficiency.

The technical scheme of the present invention is:

A spaceborne high-resolution SAR large squint Doppler dewinding method, comprising the following steps:

(1) Transform the two-dimensional spectrum data of spaceborne high-resolution SAR large squint mode;

(2) Perform support domain expansion processing on the deformed data;

(3) Two-dimensional spectrum recovery processing is performed on the data after support domain expansion.

The deformation processing is performed on the two-dimensional spectrum data of the spaceborne high-resolution SAR large squint mode, including the calculation of the deformation cyclic shift points corresponding to the data columns of the frequency point positions in each range and the edge count of the data columns of the position of the frequency points in each range. Plewit performs cyclic shift processing in two steps.

The calculation method of the deformation cyclic shift points is:

表示距离向频率点为fr_n的一列二维频谱数据。Among them, Ls represents the number of deformed cyclic shift points, F_dat_s represents the shifted data, c represents the speed of light, floor[] represents the round-down operation, V represents the effective satellite velocity, θ represents the oblique angle of view, PRF represents the system pulse repetition frequency, M is the number of received pulse echoes corresponding to a scene data, fc is the carrier frequency of the SAR system, F_dat(f _r , f _a ) indicates the two-dimensional spectrum data corresponding to a scene SAR echo, fr indicates the range frequency, and fa indicates the multiple Puller frequency,

Represents a column of two-dimensional spectral data with fr_n range frequency points.

The processing method of cyclic shift along the Doppler dimension for the data column of each range-to-frequency point position is as follows:

沿多普勒维进行循环移位处理，For a column of two-dimensional spectral data with fr_n distance frequency points

Cyclic shift processing along the Doppler dimension,

Among them, F_dat_s represents the shifted data, shft() represents the cyclic shift operation, and Ls represents the number of deformed cyclic shift points. When it is a positive number, shft() performs a positive shift, and when it is a negative number, it performs a negative shift.

Said, performing support domain expansion processing on the deformed data includes two steps of calculating the minimum number of expansion points at one end of the data and performing zero-fill expansion at the first and last ends of the deformed two-dimensional spectrum data not less than the minimum number of expansion points.

The calculation method of the minimum number of expansion points at one end of the data is:

Among them, K represents the minimum number of expansion points at one end of the data, and Fs is the AD sampling frequency.

The expansion of the support domain is realized by performing zero-padded expansion with a number of 0s not less than the minimum number of expansion points K at the beginning and end of the deformed two-dimensional spectrum data. Assuming that M is the number of received pulse echoes corresponding to a scene data, and the number of zero-padding at one end is K_m, the number of data points in the Doppler dimension after zero-padding will change from M to M+2*K_m.

Said, performing a two-dimensional spectrum recovery process on the data after the support domain expansion includes calculating the number of recovered cyclic shift points corresponding to the data columns of each distance-to-frequency point position and the Doppler along the lines of the data columns of each distance-to-frequency point position. Levy performs cyclic shift processing in two steps.

The calculation method of the number of recovered cyclic shift points is:

Here, P represents the number of cyclic shift points in the two-dimensional spectrum recovery process.

沿多普勒维进行循环移位处理，实现二维频谱恢复。其循环移位处理方式如下：Let F_dat_e denote the data after support domain expansion. For a column of expanded two-dimensional spectral data whose distance to frequency point is fr_n

Cyclic shift processing is performed along the Doppler dimension to achieve two-dimensional spectral recovery. The cyclic shift processing method is as follows:

Among them, F_dat_r represents the data after the restoration of the two-dimensional spectrum.

The advantages of the present invention compared with the prior art are:

(1) Compared with the prior art, the present invention removes the spectral Doppler wrapping of the high-resolution large squint SAR signal by performing cyclic shift deformation, support domain expansion and 2-dimensional spectrum cyclic shift recovery on the two-dimensional spectrum. phenomenon, thus solving the problems of energy leakage and false targets caused by Doppler wrapping of large squint signals;

(2) The unwrapping processing method of the present invention provides a signal unwrapping processing step flow for the signal spectrum. This process is a general signal processing process, which can be used as a preprocessing step to match with a variety of front and side imaging algorithms, and can be easily implemented in DSP, FPGA, ARM and other processors, with good versatility.

(3) The method of the present invention can simply and quickly realize the unwrapping processing under the wide-band high-resolution SAR squint, which does not involve interpolation and signal filtering calculation and other processing, avoids the calculation of echo data, and only uses signal numerical translation and data. The expansion can be realized, the processing scheme is simple and easy to implement, and has very high processing efficiency and application performance.

Description of drawings

Fig. 1 is the data processing flow chart of the present invention;

Fig. 2 is a schematic diagram of cyclic shift;

Fig. 3 is a two-dimensional spectrogram before winding under a large squint;

Fig. 4 is a two-dimensional spectrogram after spectrum deformation processing;

Figure 5 is a two-dimensional spectrogram after support domain expansion processing;

Fig. 6 is a two-dimensional spectrogram after spectrum restoration and deformation processing;

Fig. 7 is the imaging result picture when the large strabismus does not remove the winding;

Fig. 8 is a graph of the imaging result after the large squint is removed from the wrapping.

Detailed ways

The present invention will be further described in detail below in conjunction with the accompanying drawings.

The invention proposes a method for removing the Doppler wrapping phenomenon of the signal in the spaceborne high-resolution SAR large squint observation mode, and solves the problem caused by the Doppler spectrum wrapping phenomenon of the signal in the spaceborne broadband large squint observation mode. Impact. As shown in Figure 1, the present invention mainly includes three parts: two-dimensional spectrum deformation, support domain expansion and two-dimensional spectrum restoration. The specific steps are as follows:

Step 1. Transform the 2D spectrum data of spaceborne high-resolution SAR large squint mode

For the broadband SAR system, when in the large-squint observation mode, due to the broadband deformation and distortion of the two-dimensional support domain of the spectrum, the band edge exceeds the Doppler sampling bandwidth coverage in the Doppler dimension, so a two-dimensional spectrum is generated. Doppler warping phenomenon, resulting in degraded image quality.

The two-dimensional spectrum of the support domain under the strabismus view of the broadband spaceborne SAR is shown in Figure 3. It can be seen from the figure that in the case of broadband deformation and distortion of the spectrum, the two-dimensional spectrum exceeds the sampling frequency band, the spectrum in the upper left corner is wrapped to the upper right corner, the spectrum in the lower right corner is wrapped to the lower left corner, and the spectrum appears. severe fracture.

Aiming at this phenomenon, the two-dimensional spectral data is deformed first, so that the spectral Doppler support domain after the deformation is within the Doppler sampling bandwidth.

For the digital signal sampling SAR system, let V represent the effective speed of the satellite, θ represents the oblique angle of view, PRF represents the system pulse repetition frequency, M represents the number of received echo pulses corresponding to a scene data, Fs represents the AD sampling frequency, and N represents the The number of AD sampling points corresponding to an echo pulse, f _c is the carrier frequency of the SAR system, F_dat(f _r , f _a ) is the two-dimensional spectrum data corresponding to a scene SAR echo, f _r represents the range frequency, and f _a represents the multiple Puller frequency.

The deformation processing is performed on the two-dimensional spectrum data of the spaceborne high-resolution SAR large squint mode, including the calculation of the deformation cyclic shift points corresponding to the data columns of the frequency point positions in each range and the edge count of the data columns of the position of the frequency points in each range. Plewit performs cyclic shift processing in two steps.

The specific deformation method of two-dimensional spectral data is as follows:

沿多普勒维进行循环移位处理，利用如下公式计算变形循环移位点数Ls：For a column of two-dimensional spectral data with fr_n distance frequency points

Perform cyclic shift processing along the Doppler dimension, and use the following formula to calculate the number of deformed cyclic shift points Ls:

The way to deform the data column using the deformed cyclic shift points is as follows:

Among them, F_dat_s represents the shifted data, c represents the speed of light, shft() represents the cyclic shift operation, floor[] represents the round-down operation, and Ls represents the number of cyclic shift points. When it is a positive number, shft() performs forward direction Shift, negative shift if negative.

Figure 2 shows the transformation process of a data column of spectral data. The shaded part of the two-dimensional spectrum in Figure 2(a) represents a data column distributed along the Doppler dimension. Taking Ls=2 and M=7 as an example, Fig. 2(b) shows a schematic diagram before and after the data column is moved.

Figure 4 shows the spectral shape after performing cyclic shift processing on all data columns of the two-dimensional data to achieve spectral deformation.

Step 2. Perform support domain expansion processing on the deformed data

The two ends of the data after the two-dimensional spectrum deformation are filled with zeros to realize the expansion of the support domain. Among them, the number of zero-padding at one end should not be less than the following number:

Among them, K represents the lower limit of the number of zeros at one end. During processing, no less than K number of 0s are added at the beginning and end of the Doppler dimension of the data to realize data expansion. Assuming that the number of front-end zeros is K_m1 and the number of back-end zeros is K_m2, the number of data points in the Doppler dimension after zero-filling will change from M to M+K_m1+K_m2.

The 2D spectrogram after support domain expansion is shown in Figure 5.

Step 3: Perform two-dimensional spectrum recovery processing on the expanded data of the support domain to realize the Doppler domain unwrapping of the two-dimensional spectral data

Performing two-dimensional spectrum recovery processing on the data after the expansion of the support domain includes calculating the number of recovered cyclic shift points corresponding to the data columns of each range-to-frequency point position and circulating the data columns of each range-to-frequency point position along the Doppler dimension. The shift process is two steps.

沿多普勒维进行循环移位处理，实现二维频谱恢复。其恢复循环移位点数Lr如下：Specifically, let F_dat_e denote the data after support domain expansion. Similarly, for the two-dimensional spectral data after the expansion of a column of support domains whose distance to the frequency point is fr_n

Cyclic shift processing is performed along the Doppler dimension to achieve two-dimensional spectral recovery. The recovery cyclic shift point Lr is as follows:

The way to deform the data column by recovering the cyclic shift points is as follows:

表示距离向频率点为fr_n的一列支撑域扩展后的二维频谱数据恢复后的数据，shft()表示循环移位操作。in,

Represents the restored data from the two-dimensional spectral data extended to a column of support domains whose distance to the frequency point is fr_n, and shft() represents a cyclic shift operation.

After the two-dimensional spectrum recovery process, the broadband large squint two-dimensional spectrogram of the two-dimensional spectrum unwrapping is shown in Figure 6. It can be seen from the figure that the shape of the spectrum has been effectively restored after processing. Comparing the coiled spectrum in Figure 3, it can be found that the spectrum distribution after de-wrapping is continuous, and the coiling and fracture phenomena have been effectively removed. .

The dewinding method can be used as a general preprocessing algorithm, and can be integrated with existing imaging algorithms without modifying the imaging algorithm. For the large strabismus observation data without 2D spectral unwrapping processing, the imaging processing result is shown in Figure 7. It can be seen from the figure that there are energy leakage and blurred and defocused targets on both sides of the target. After dewinding preprocessing by this method, the existing imaging algorithm is used to perform imaging focusing processing on the dewinding data, and the obtained imaging result is shown in FIG. 8 . Comparing Fig. 7, it can be seen that the energy leakage interference on both sides of the target disappears, indicating that this method effectively realizes the removal of spectrum wrapping and greatly improves the imaging results.

The method of the invention is aimed at the high-resolution large-squint observation mode of the SAR satellite, and realizes the removal of the two-dimensional spectral Doppler domain wrapping phenomenon by processing such as rapid deformation of the two-dimensional spectrum, expansion of the support domain, and rapid restoration of the two-dimensional spectrum. Energy leakage and a large number of false targets due to spectral wrapping in the Doppler domain. The dewinding method of the present invention realizes spectrum processing by utilizing the cyclic shift operation, avoids the calculation of echo data, and has very high processing efficiency and application efficiency.

The parts not described in detail in the present invention belong to the common knowledge of those skilled in the art.

Claims (7)

1. A satellite-borne high-resolution SAR high squint Doppler deconvolution method is characterized by comprising the following steps:

the method comprises the following steps: carrying out deformation processing on the two-dimensional spectrum data of the satellite-borne high-resolution SAR large squint mode;

dividing the two-dimensional frequency spectrum data into data columns at each distance direction frequency point position, calculating the deformation cyclic shift point number corresponding to each data column, and performing deformation processing on each data column by using the deformation cyclic shift point number;

step two: carrying out support domain expansion processing on the deformed data;

step three: performing two-dimensional spectrum recovery processing on the data after the support domain expansion to realize the unwinding of the Doppler domain of the two-dimensional spectrum data;

the method for carrying out two-dimensional spectrum recovery processing on the data after the support domain expansion comprises the following steps: dividing the data after the support domain expansion into data columns at the positions of the frequency points in each distance direction, calculating the number of recovery cyclic shift points corresponding to each data column, and performing deformation processing on each data column by using the number of the recovery cyclic shift points.

2. The spaceborne high-resolution SAR high squint Doppler deconvolution method according to claim 1, characterized in that: for a column of two-dimensional spectrum data with distance to frequency point fr _ n

Calculating the number Ls of deformation cyclic shift points by using the following formula:

wherein c represents a light velocity, floor 2]Representing a rounding-down operation, V representing the satellite effective velocity, theta representing the squint angle, PRF representing the system pulse repetition frequency, f_cFor SAR system carrier frequency, F _ dat (F)_r，f_a) Representing two-dimensional spectral data corresponding to a scene of SAR echoes, f_rRepresenting the range frequency, f_aThe Doppler frequency is shown, and M is the number of received pulse echoes corresponding to one scene of data.

3. The spaceborne high-resolution SAR high squint Doppler deconvolution method according to claim 2, characterized in that: the method for performing deformation processing on the data column by using the deformation cyclic shift point number is as follows:

wherein,

represents data shifted from a row of two-dimensional spectrum data with a frequency point fr _ n, shft () represents a cyclic shift operation, and shft () shifts positively when Ls is a positive number and shifts negatively when Ls is a negative number.

4. The spaceborne high-resolution SAR high squint Doppler deconvolution method according to claim 1, characterized in that: in the second step, the method for performing support domain expansion processing on the deformed data comprises the following steps:

(2.1) calculating the minimum expansion point number at one end of the data;

and (2.2) respectively performing zero filling expansion not less than the minimum expansion point number at the head end and the tail end of the two-dimensional frequency spectrum data subjected to the deformation processing in the step one, and realizing the expansion processing of the data support domain.

5. The spaceborne high-resolution SAR high squint Doppler deconvolution method according to claim 4, characterized in that: in the step (2.1), the minimum expansion point number K at one end of the data is calculated by using the following formula:

wherein Fs is AD sampling frequency, c is light speed, floor [ ] is rounding-down operation, V is satellite effective speed, θ is squint angle, PRF is system pulse repetition frequency, and M is the number of received pulse echoes corresponding to a scene data.

6. The spaceborne high-resolution SAR high squint Doppler deconvolution method according to claim 1, characterized in that: two-dimensional spectrum data after expanding a column of support domains with distance to frequency point fr _ n

The number of recovery cyclic shift points Lr is calculated using the following formula:

wherein c represents a light velocity, floor 2]Representing a rounding-down operation, V representing the satellite effective velocity, theta representing the squint angle, PRF representing the system pulse repetition frequency, f_cFor SAR system carrier frequency, f_rRepresenting the range frequency, f_aThe Doppler frequency is shown, and M is the number of received pulse echoes corresponding to one scene of data.

7. The spaceborne high-resolution SAR high squint Doppler deconvolution method according to claim 6, characterized in that: the data column is deformed by the recovery cyclic shift point number as follows:

wherein,

representing two dimensions after extension of the distance to a column of support fields with frequency points fr _ nFor the spectrum data recovered data, shft () represents a cyclic shift operation.

Images